IMCCE: Reaching for the Stars
French institute turns to Intel® Xeon® processor 7500 series to accelerate their research concerning the dynamics of the planetary systems
The Institut de Mécanique Céleste et de Calcul des Ephémérides (IMCCE) researches celestial mechanics and the dynamics and astrometry of solar system objects. It can trace its origins back to 1795 when it produced ephemeris charts in almanacs. These charts show the positions of solar system celestial bodies on given dates and in regular sequence. Today the IMCCE still produces these charts, which are used by astronomers all over the world. In 1988 the IMCCE helped develop an application called TRIP*–www.imcce.fr/trip–an interactive computer algebra system specially adapted to celestial mechanics, which is still in use today.
Challenges
• Research imperatives: IMCCE wanted to increase the speed at which it carried out astronomical research
• Software optimization: IMCCE uses TRIP, software designed specifically for celestial mechanics, which needed to be optimized for enhanced performance on servers with more than 32 cores
• Memory needs: TRIP is a memory-intensive application, so besides greater CPU performance, IMCCE needed a system that provided large memory caches to ensure top performance
• Solutions
• Benchmarking: IMCCE benchmarked the Intel® Xeon® processor 7500 series against its existing servers powered by the Intel® Xeon® processor 5500 and 5600 series
• Tuning tools: It also utilized Intel® C++ Compiler and Intel® vTune™ Performance Analyzer to optimize the TRIP code and trace any potential performance bottlenecks
• Helping hand: The benchmarking was carried out on an HP ProLiant* DL980 G7 server, which HP fine-tuned to ensure optimal performance
Impact
• Seven times faster: The HP Intel Xeon processor 7500 series-powered system provided up to seven times faster performance than IMCCE’s existing system
• Memory depth: Up to 24MB of last level cache for each processor ensured top performance for the memory-intensive TRIP application
• Built-in flexibility: The HP ProLiant* DL980 G7 server has good socket scalability and can deliver up to eight cores per socket, as such it is easily expandable so further memory can be added
Read the full IMCCE: Reaching for the Stars Case Study.
938KB
We are sorry, This PDF is available in download format only
IMCCE: Reaching for the Stars
French institute turns to Intel® Xeon® processor 7500 series to accelerate their research concerning the dynamics of the planetary systems
The Institut de Mécanique Céleste et de Calcul des Ephémérides (IMCCE) researches celestial mechanics and the dynamics and astrometry of solar system objects. It can trace its origins back to 1795 when it produced ephemeris charts in almanacs. These charts show the positions of solar system celestial bodies on given dates and in regular sequence. Today the IMCCE still produces these charts, which are used by astronomers all over the world. In 1988 the IMCCE helped develop an application called TRIP*–www.imcce.fr/trip–an interactive computer algebra system specially adapted to celestial mechanics, which is still in use today.
Challenges
• Research imperatives: IMCCE wanted to increase the speed at which it carried out astronomical research
• Software optimization: IMCCE uses TRIP, software designed specifically for celestial mechanics, which needed to be optimized for enhanced performance on servers with more than 32 cores
• Memory needs: TRIP is a memory-intensive application, so besides greater CPU performance, IMCCE needed a system that provided large memory caches to ensure top performance
• Solutions
• Benchmarking: IMCCE benchmarked the Intel® Xeon® processor 7500 series against its existing servers powered by the Intel® Xeon® processor 5500 and 5600 series
• Tuning tools: It also utilized Intel® C++ Compiler and Intel® vTune™ Performance Analyzer to optimize the TRIP code and trace any potential performance bottlenecks
• Helping hand: The benchmarking was carried out on an HP ProLiant* DL980 G7 server, which HP fine-tuned to ensure optimal performance
Impact
• Seven times faster: The HP Intel Xeon processor 7500 series-powered system provided up to seven times faster performance than IMCCE’s existing system
• Memory depth: Up to 24MB of last level cache for each processor ensured top performance for the memory-intensive TRIP application
• Built-in flexibility: The HP ProLiant* DL980 G7 server has good socket scalability and can deliver up to eight cores per socket, as such it is easily expandable so further memory can be added
Read the full IMCCE: Reaching for the Stars Case Study.
